From a discussion of fusion reactor designs based on today's well-established experience gained in the operation of large tokamaks, it is concluded that such reactors are economically not attractive. The physics involved in the various options for concept improvement is described and the main experimental results of current research are shortly reviewed.